Automotive production control, using thermal vision systems-A passive thermal imagery for process control

M. Omar, Y. Zhou

    Research output: Contribution to journalArticlepeer-review


    This work presents new generation of online inspection systems, developed and deployed in automotive assembly plants, to provide 100%, online evaluation of automotive processes and subassemblies. The presented systems are thermal-vision based, with capability to provide two dimensional, spatial and temporal temperature acquisitions, and are equipped with novel processing routines for data collection, conditioning and processing. Specific systems are presented in the stamping presses, paint curing practices, with inspection schemes designed to provide an efficient manufacturing sequence, through (1) pro-active diagnosis routines to monitor the process performance, while inspecting the product, (2) focusing on the error detection rather than defect detection routines, and (3) utilizing closed loop, feedback communications. This work will also present the processing computations that combine known image processing tools with inverse conduction computations, to correlate the process, product health parameters, to acquired scans temperature readings. Presented work will include two implementations of proposed thermographic systems; the automated detection of splits for body in white BIW sub-assemblies in automotive stamping presses, in addition to stamping die temperature control. Second example is a thermal vision system implemented, to evaluate the protective coating coverage and its curing uniformity for automotive, steel fuel tanks.

    Original languageBritish English
    JournalSAE Technical Papers
    StatePublished - 2008


    • automotive assembly lines
    • Non-contact split detection
    • quality control
    • Thermal machine vision


    Dive into the research topics of 'Automotive production control, using thermal vision systems-A passive thermal imagery for process control'. Together they form a unique fingerprint.

    Cite this